How much does it cost to start an online store?

The budget you wrote down has three lines on it: a platform subscription, a domain, and "some ads". The bill that actually arrives has nine lines, and the two largest are usually the ones you left off.

So, how much does it cost to start an online store? A few hundred dollars if you sell something you make on a computer. One to three thousand if you buy stock before you have buyers. Almost nothing beyond the software if you already have an audience that trusts you. The spread is that wide because most of the cost is not the store.

The nine lines that are actually on the bill

Seven of those are easy to look up. The eighth is a reserve people decline to hold. The ninth deserves its own section, because it is the difference between a budget that survives month one and a budget that does not.

The line nobody budgets: the marketing that produces nothing

Your first marketing money does not buy sales. It buys information: which audience responds, which image stops the scroll, which price makes people hesitate, which product people click but do not buy. That information is genuinely valuable and it is genuinely not free, and because it does not show up as revenue, almost nobody puts it in the spreadsheet.

Three honest ways to pay for it.

• Pay cash. Set a number you can lose entirely and treat it as tuition. If a week of ads returns nothing but tells you the product page loses people at the shipping line, the money worked.
• Pay in time. Posting, messaging people individually, showing up where your buyers already are. Slower, usually cheaper in dollars, and not free — it costs your evenings, which is the resource you have least of.
• Borrow someone's audience. A collaboration, a newsletter swap, a marketplace listing. Cheapest of the three and the reason a marketplace listing is worth considering even when you have your own store.

Software can lower the cost of producing marketing. The mAutomate AI Marketing Suite drafts posts, campaigns and ad copy and schedules them, with nothing publishing until you approve it. What it cannot lower is the cost of learning — the first version of anything is a guess, whether a person or a model wrote it, and the market is the only thing that settles it.

Scenario one: digital products

A pattern, a preset pack, a course, a template. No stock, no shipping, no breakage. The cheapest real store you can run.

The honest part of that table is the word "nothing" in the marketing row. The dishonest version of this post would show $41 and call it a launch.

Two things make the digital number move. First, whether you can shoot or design your own cover art — the gap between $0 and $150 is entirely whether you own that skill. Second, whether refunds are generous; digital refunds are cheap to grant and easy to abuse, so decide the policy before launch rather than during the first awkward email. If you have not picked what to sell yet, the selection problem is a separate one.

Scenario two: print on demand

Someone else holds the stock, prints on order and ships. You pay a unit cost out of each sale instead of buying inventory up front, which is why this is the model most people should start with and the model with the worst margins.

The sample order is the line people cut and should not. You are putting your name on a garment you have never touched, printed by a supplier whose color accuracy you are guessing at. Three samples cost less than one wave of refunds and one bad review, and the photographs you take of them will outperform the supplier's mockups.

What this table hides: per-unit economics. Print on demand turns a one-off cost into a permanent one, so a product that works brilliantly at ten orders a month can be irritatingly thin at three hundred. Run the pricing arithmetic before you run the ads, and read the setup sequence before you order samples.

Scenario three: buying physical stock

You buy inventory, you store it, you pack it, you ship it. Best margins of the three, worst cash position, and the only one of the three where you can be properly wrong — wrong size mix, wrong color, wrong quantity.

Two notes on the stock line. It is a range because it is the one number you choose, and the common error is choosing it by what you can afford rather than by what you can photograph, store and ship from where you currently live. Boxes arrive in a hallway. The hallway has a size.

The returns reserve looks like paranoia until the first mug arrives in three pieces. Five percent is an illustrative placeholder; your real figure depends on how fragile the product is and how far it travels, and you only learn it by shipping.

The only real prices in this post

Billed monthly; six-month and yearly terms are discounted. Every plan starts with a 14-day free trial and no card, so the software line of your month-one budget can genuinely be zero while you find out whether the rest of the plan survives contact with reality. Full detail is on the pricing page.

Where this arithmetic breaks

Your own time is the largest cost and it is not in any table. A launch that takes you sixty evenings cost sixty evenings. If you would rather not spend forty of them choosing fonts, that is a real argument for software that drafts the storefront for you — but it is a time argument, not a money one, and it should be made as one.

Country changes almost every line. Processing fees, postage, domain pricing, what a photographer charges, whether you can even get the supplier to ship to you. A cost post written for one country and read in another is off by a multiple, not a rounding.

Cheap launches are not automatically cheaper. A $200 launch that sells nothing for five months cost $200 plus five months. A $2,000 launch that finds a buyer in week three is the better deal and looks worse in the spreadsheet. Cost per launch is the wrong metric; cost per unit of learning is closer.

Nothing here is a forecast. These are shapes of arithmetic, not predictions about your store. Anyone who tells you what your store will cost without knowing what you sell, where you live and who you can already reach is guessing with more confidence than the data supports.

Redo it with your own figures

Six lines. Fill them in from quotes, not from blog posts, including this one.

1. Platform, from the table above. Real number.
2. Domain, from whoever you buy it from, for the first year.
3. The thing you sell: cost to make one, or cost of the first order, or unit cost per sale.
4. Images: what it costs you to get one clean photograph of each product.
5. Shipping: mailer plus label plus filler, per order.
6. The number you are willing to lose finding out whether anyone wants it.

Add lines one through five and you have your floor. Line six is the number that decides whether the store gets a fair trial, and if you cannot name it yet, name it before you launch rather than after. If you want the sequence rather than the arithmetic, the launch checklist is here, and the platform fee side of the bill is here.